Media Gateway Controller (MGC) redundancy concept is employed in present day communication networks. MGC redundancy concept is based on a one-plus-one concept or active-active concept, where both the MGC's are maintained in active state for operation. In cases, wherein one of the MGC fails or may be taken out for maintenance reasons, the other MGC of the pair takes over the control of Media Gateway (MG) served by the failed MGC. The concept is based on static allocation of MG to the MGC of a pair. Each MG has a preliminary link connected to one MGC (primary MGC) and a secondary link connected to another MGC (secondary MGC). During manual maintenance of MGC, MG served by the MGC may be directed to secondary MGC by sending a hand-off request.
The drawbacks involved in the above mentioned hand-off mechanism is that, if a restart occurs on the MG after the MG is successfully registered with the secondary MGC, the MG tries to connect to the primary MGC instead of secondary MGC. Further, if the maintenance is not yet started on the primary MGC, the primary MGC accepts the connection request from the MG without the knowledge of network operator. As a result, the MG is not served when the MGC goes for maintenance. On the other hand, when there is an internal failure in the processor handling the concerned MG, the connection between the MG and secondary MGC is lost. The MG may again re-register with the primary MGC by sending a service request to establish the connection. But the primary may not be able to serve the MG as it is under maintenance. Due to these problems, all calls running on the concerned MG will be affected. Also, the network operator is not aware that a MG is actually registered to the MGC under maintenance.
MG and MGC are separated by a long distance, and actions done at the MG side are not known at the MGC level. Data related to all MGs are configured before hand on the MGC. If a new MG is to be serviced by a primary MGC, which is to be under maintenance, the service request sent from the MG is accepted by the primary MGC without the knowledge of the network operator. Because of the above reasons, calls running on the concerned MG are affected and abrupt release of calls takes place. Also, there is no standardization aspect in existing hand-off mechanisms.